Travel companies are putting their AI money into the back office first: in Sabre's survey of more than 140 airline, agency and hotel executives, 70% are investing in workforce tools and workflows. Only 17% are reducing AI technology spend, and the people paying have decided the first return sits inside the office.
Key facts
- 70% of the executives are investing in workforce tools and workflows — Sabre, Compass: Navigating the Fog.
- 17% are reducing AI technology spend, while 41% are protecting their core roadmap and 27% are accelerating — Sabre.
- 73% of travel agencies are confident about the next 12 months, against 66% of all the executives — Sabre.
- 38.1% of the more than 115,000 priced options quoted to agencies could be tied to a named hotel — the author's operations archive.
The news. Sabre's study "Compass: Navigating the Fog" surveyed more than 140 airline, agency and hotel executives in June. Only 17% are reducing AI technology spend. 41% are protecting their core roadmap, 27% are accelerating and 15% are moving spend towards cost reduction and efficiency tools. On where the money is going, 70% are investing in workforce tools and workflows. Three-quarters ranked geopolitical instability as their top concern; for agencies, the study says, it shows up "as unpredictable booking lead times and shakier consumer commitment". 66% of the executives say they are confident about the next 12 months, and among travel agencies the figure is 73%.
Sabre's chief operating officer, Shawn Williams, framed it this way: "The companies that come out ahead won't be the ones that waited for the fog to lift."
Why do workflows come first in travel AI budgets?
Budgets under pressure go where they pay back soonest, and inside a travel business that is rarely the showpiece a traveller sees. It is the work behind it: the requests, the quotes, the changes, the follow-ups. Seven in ten of these executives are putting money there.
That is a more useful signal than any launch announcement. It says the people paying for AI in travel have decided the first return is inside the office.
How messy are a desk's real inputs?
Workflow tools are only as good as what they are fed, and a desk's real inputs are not tidy. On the desks I ran, only 38.1% of the more than 115,000 priced options quoted to agencies could be tied to a named hotel when I measured them. Anything built on hotel data starts from that share.
That is the case for workflow money and the warning attached to it. The work is already there, in volume. Turning it into data a system can use is work too, and it is the part a demo never shows.
Why does it matter for the agency?
Agencies in this sample are more confident than the industry around them. Spend that confidence on the back office before the front. Start with the requests you already handle every day, and before you buy a system, measure how much of your own data a system can actually read.